GENESIS: Documentation

Related Documentation:

Functional User Specification for the G-Tube

This functional user specification for the GENESIS GUI (G-Tube) is based on the four steps of the Publication Workflow (see Fig. 1). This specification covers single cell models.


PIC

Figure 1: Publication Workflow: The four steps of the Publication Workflow. The first step of the Publication Workflow is provided by the User Workflow (see Fig. 2).


  1. User Workflow:
    Selecting the “User Workflow” tab presents the User Workflow panel that contains the following selectable menu items (see Fig. 2).


    PIC
    Figure 2: User Workflow: The five steps of the User Workflow. The completed User Workflow is embedded in the first step in the Publication Workflow (see Fig. 1).


    1. Construct Model:
      1. Import Model:
        Load and open a pre-existing project environment in the GUI. Default: NDF format. Other loadable file formats include: NeuroML, 9ML, or from a GENESIS 2 model imported via the NS-SLI module.
      2. Create Model:
        1. Create single compartment model:
          • Create compartment.
          • Define passive membrane parameters.
        2. Create multi-compartment model:
          • Import cell morphology.
          • Define passive membrane parameters for soma and dendrites. Default: Somatic and dendritic parameters identical.
        3. Populate model:
          • Add channel types from channel model library.
          • Define ˉg for each channel type.
          • Define 1-D distribution of channel densities for each channel type included in multicompartment model. Choice of:
            • Uniform: Default for dendrites and soma. Can be set independently for soma and dendrites.
            • Gamma: Dendritic density defined by absolute ortho- or antidromic distance from soma along single dendrites.
      3. Modify model parameter values:
      4. Model summary:
        Generate tables containing passive membrane properties and channel, synaptic, and gap-junction parameter values (if present).
      5. Save tables: Generated during model checking step of User Workflow.
        • Black: Default values.
        • Green: User modified values.
        • Red: Problematic values, e.g. out of expected range.
      6. Annotate tables:
      7. Explore model:
        1. Generate list of model compartments: Selecting one or more compartments from this list generates a list of properties, components, and parameter values for each selected compartment.
        2. Selected compartments: Map to 3-D scalable and rotatable cell morphology.
        3. Generate list of model components: Selecting a model component generates a list of compartments that contain the given component. Location of selected components map to 3-D scalable and rotatable cell morphology.
      8. Annotate model:
        Record source of any changes made to published parameters (the default parameter values of the model). This should include whether the source is one of the following:
        • Citation for published source of new values.
        • Obtained from simulations.
      9. Save model: Choice of NDF, NeuroML or 9ML file formats. Model can be saved in one or more formats. Default format: NDF.
    2. Design Experiment:
      1. Choose inputs:
        1. Set stimulus type: One of:
          • Current injection.
          • Current clamp.
          • Voltage clamp.
          • Synaptic stimulation.
        2. Set specific parameter values of stimulus protocol:
          • Current injection: Set magnitude and duration.
          • Current clamp: Set magnitude and duration.
          • Voltage clamp: Set holding potential and duration.
          • Synaptic stimulation:
            • Dendritic location.
            • Number of synapses.
            • Frequency.
            • Temporal distribution of impulses: uniform, poisson, gamma.
        3. Choose outputs:
          • Select and set parameter values to be saved.
        4. Choose output resolution:
        5. Save experimental design: Automatically saved as part of SSP file.
    3. Run Simulation:
      1. Set and check runtime options:
        • Update time step duration: fixed or variable.
        • Simulation duration.
        • Visualize runtime parameters.
      2. Run simulation:
      3. Save model state: This can be used to eliminate time required for model startup.
      4. Reset simulation: Set solver variables to values given by one of:
        • Default: Initial state of last model saved.
        • Specified previously saved model state.
    4. Output:
      1. Construct draft atoms: For example see Fig. 3.
        • Internally: Via g3plot and/or the Studio.
        • Externally: Via Matlab, xmgrace, Mathematica, etc.
      2. Import externally generated figures, graphs and tables:
      3. Check simulation output: Does expected output exist.
      4. Check validity of results: Does output make sense.
      5. Analyze output: Statistical analysis.
      6. Annotate draft atoms:.


      PIC
      Figure 3: Figure Molecule: Example relationship between publication atoms and a (figure) molecule.



      PIC
      Figure 4: Atom Pipeline: Technical specification of relationship between publication atoms and the Publication Workflow.


    5. Iterators:
      1. Batch file construction:
        • Manual: From text editor generate SSP file variants..
        • Automated: Use scripts to automatically generate SSP file variants.
      2. Dynamic Clamp: RTXI.
      3. Note: During completion of the User Workflow a numbered list is constructed of all “Draft” atoms (see Fig. 4).
      4. Select Draft Atoms for inclusion in Manuscript: The components of a project to be included in a publication are selected from draft atoms generated during completion of the User-Workflow. In the content selection step (following), these atoms become ”Manuscript” atoms.
  2. Content Selection:
    1. Gather selected manuscript atoms:
    2. Organize selected manuscript atoms: The list of atoms chosen by an investigator for publication can be ordered.
  3. Automated Model Validation:
    Ensure lineage starting point of the model employed for simulations is a previously published source model.
    1. Model verification: Performed on
      • Morphology.
      • Cable discretization.
      • Equations for different membrane and synaptic channel types and their kinetics.
      • Model parameter values.
      • Dynamic response to specific current injection and voltage clamp protocols.
    2. Test robustness of parameter values: Assess fragility of new or changed model parameters.
    3. Verify parameter values: Determine ‘normality of new or changed parameters used by model.
    4. Check publication atoms: Numerical values cited in publication atoms are compared with model values.
    5. Verify figures: Check all figures are generated by the same model.
    6. Identify changes to base model: Compare project model to base model.
      • Check morphology.
      • Check parameter values.
    7. Identify source of changed values: Are changed values from new experimental data or simulations.
  4. Peer Review/Publication:
    1. Automated validation report: Provides technical summary.
      • Morphology.
      • Parameters.
      • Equations.
      • Figures.
      • Tables.
      • References.
    2. Human review:
      • Narrative.
      • Model.
      • Model extensions.
      • Novelty.
      • Significance.